MR002-0010
Global Sensitivity Analysis for U(VI) Transport in a Generic Deep Geologic Repository
Monday, 14 December 2020
Poster
Dinara Ermakova1, Haruko M Wainwright2, Liange Zheng3, Ian Shirley2, Hannah Lu4 and Daniel Tartakovsky5, (1)University of California Berkeley, Nuclear Engineering, Berkeley, CA, United States, (2)Lawrence Berkeley National Laboratory, Berkeley, CA, United States, (3)Lawrence Berkeley National Laboratory, Energy Geosciences Division, Berkeley, CA, United States, (4)Stanford University, Energy Resources Engineering, Stanford, CA, United States, (5)Stanford University, Stanford, United States
Abstract:
The performance assessment (PA) for nuclear waste disposal has been developed and improved over the past decades to compute the radiological risk. PA involves the evaluation of multiple components along with the radionuclide transport (RT) pathway to the biosphere, as well as considers a variety of processes that affect radionuclide transport, the ultimate measure of repository performance. Recent research has been focused on complex interactions between host rock and buffers such as coupled Thermal, Hydrological, and Chemical (THC) processes, and their associated impacts on RT. There are many parameters that potentially affect these processes, including groundwater content, flow, waste form composition. The uncertainty associated with these processes is not accounted for in the overall PA and poses a great challenge for the integration of multiple coupled processes models into the global PA model. Without this integration, it is impossible to identify important parameters for reducing the overall uncertainty (OU).
In this study, numerical reactive transport simulations and sensitivity analyses were performed to assess geochemical processes controlling the migration and retardation of U(VI) from a generic nuclear waste repository system. The particular focus is to use the THC model for computing apparent Kd as a function of time and space that can be plugged into the PA model. We simulated the THC process within a buffer-argillite system using TOUGHREACT and computed Kd at each grid block over time. The spatial averaging strategy was considered to determine the Kd value representing the entire buffer as one grid block in the PA model.
Sensitivity analysis was performed by integrating the TOUGHREACT code with the PEST protocol of ITOUGH2. Predicted Kd is sensitive to geochemical parameters, and that the sensitivity of these predicted concentrations evolves in space and time. During the early stage predicted Kd in constant and the most influential parameters are adsorption sites density on smectite, pH, Ca2+ concentration in the aqueous phase, and smectite volume fraction. The Kd and major geochemical changes occur after 10 000 after the waste canister placement.
